Hybrid baseline localization for portable AUV navigation

In this paper we present experimental and simulated results for an improvement upon the moving short baseline (MSBL) method of navigation for a fleet of autonomous underwater vehicles (AUVs). With the goal of maintaining system simplicity and portability, the hybrid baseline (HBL) approach utilizes one or both of the objective ship-mounted MSBL transponders in conjunction with a single off-path drifting buoy-mounted transponder intended to resolve problems observed in previous MSBL navigational work. The states of the ship and buoy are estimated using extended Kalman filters (EKFs) that are updated with state broadcasts from each source. Because the system uses a synchronous communication cycle, the AUVs are also able to calculate ranges with each broadcast; with the addition of onboard sensor data, the AUVs are able to use these ranges to estimate their own state using a third EKF. Simulation and field test results indicate that with respect to previous MSBL results, the perpendicular-to-path range data from the buoy allows the EKF to more quickly reduce the initialization error in the AUV position and maintain a lower overall position error.